Saving Groundwater from Stubborn Pollutants

Spray aeration vapor extraction systems and other innovative technologies are successfully treating recalcitrant compounds

Groundwater contamination has been an area of environmental concern for decades. Underground storage tank and process line leaks, product spills and intentional dumping have been the greatest contributors to the problem. Over the years, many in-situ and ex-situ treatment methodologies have been developed and implemented to remediate impacted sites. The most common methods employed have been based on the "pump-and-treat" concept, where contaminated groundwater is removed from the aquifer via mechanical means and is treated by external processes.

Historically, these processes include adsorption, transfer of contaminant into vapor phase (air stripping), ultraviolet treatment and chemical oxidation. While these processes have generally been proven to be effective, they can be quite costly, owing to regulatory monitoring requirements for effluent discharge, vapor phase treatment in non-attainment areas and operating and maintenance costs, such as media replacement/regeneration, power, supplemental process fuels, etc.

In-situ treatments have also been employed where practicable and have generally been based upon applying the operating principles of the processes described above and modifying the process to perform remedial activities without removal and disposal of the contaminated water.

Early on, the contaminants of concern were petroleum and chlorinated hydrocarbons, including BTEX (benzene, toluene, ethylbenzene and xylenes), PCE (perchloroethylene) and TCE (trichloroethylene). In more recent times, additional contaminants have been identified for remedial action such as MTBE (methyl tertiary butyl ether), TBA (tertiary butyl alcohol), 1, 4-dioxane and perchlorates. Unfortunately, many of these more recently identified contaminants fall into the "recalcitrant" category as their chemical composition and general affinity for water makes their removal and treatment much more challenging and costly than previously identified contaminants. Further, traditional treatment methodologies have generally proven to be less effective when applied to recalcitrant compounds.

The environmental industry has been diligently working to address the new challenges of remediating recalcitrant compounds. Two companies have developed enhancements to proven treatment processes and have further combined discrete technologies in a synergistic approach to reduce capital equipment costs and improve process destruction performance for recalcitrant compounds. These treatment processes are as follows:

  • Spray aeration vapor extraction system and hollow-fiber membrane technology. Remediation Service International (RSI), a division of Innovative Environmental Solutions LLC located in Ventura, Calif., has enhanced conventional air stripping technology by the addition of heat, vacuum and spray aeration and has developed the patented S.A.V.E.™ (spray aeration vapor extraction) system. Additionally, RSI has developed a hollow-fiber membrane technology (patent pending) that may be employed on a stand-alone basis or in conjunction with its S.A.V.E. system. Both of these technologies are effective in treating groundwater that has been contaminated with BTEX and MTBE, and are also offered as optional modules for use in conjunction with RSI's internal combustion engine-based (ICE) vapor extraction treatment systems. Authors Keller and Tokat of the University of California, Santa Barbara, have developed a predictive model for contaminant removal efficiencies of these systems, which is described in detail in this article.
  • Advanced oxidation process technologies. Applied Process Technology Inc., based in Pleasant Hill, Calif., has developed two advanced oxidation process (AOP) technologies that effectively destroy recalcitrant groundwater contaminants such as MTBE, TBA and 1,4-dioxane. The first, called HiPOx™, is a patented technology designed for pump-and-treat applications. The other, called Vadozone™ (patent pending), is an in-situ treatment solution. Several case studies are presented for these processes that address the treatment of various recalcitrant compounds.

Remediation Service International's Remediation Techniques
RSI's S.A.V.E. system is an ex-situ technique used for remediation of contaminated groundwater. Removal is accomplished by passing the contaminated water through a counter current air stream in a spray aeration tank maintained under high vacuum. The water is heated using a liquid heat exchanger and then sprayed at approximately 25 pounds per square inch (psi) through seven nozzles using an 80 gallons per minute- (gpm) rated centrifugal pump prior to discharge. Because of the spraying effect, small drops of water are formed in the order of 1 millimeter to 2 millimeters in diameter. These droplets help transfer contaminants, such as BTEX and MTBE, from liquid phase to gas phase. This process is an enhanced form of air stripping, since low pressure and high temperature are used to enhance mass transfer. After the removal of contaminants from the water to the gas phase, these gases are collected and forwarded to the engine combustion system, where contaminants are destroyed into non-hazardous forms.

In addition to the S.A.V.E. system, hydrophobic hollow fiber membranes (HFM) (Figure 1) can be used in order to transfer water-soluble contaminants from liquid to gas phase by physical means. An HFM contains fibers made of synthetic porous material, such as polypropylene, which are hydrophobic in nature. While the contaminated water is passing through the membrane, vacuum is also provided continuously. This helps to keep a high concentration gradient between the liquid phase and the gas phase to achieve removal of contaminants. The membrane provides a support medium to help liquid and gas phase come into contact, where mass transfer occurs. This process results in the separation of water from volatile and semi-volatile gases.



Figure 1. Hollow Fiber Membrane Details (Copyright RSI 2003)

In this system, S.A.V.E. and HFM modules are combined in order to increase the efficiency of contaminant removal and meet the goal effluent concentrations. Contaminated water is passed through the S.A.V.E. system first and then the HFM system, where four HFM modules are placed in parallel. Effluent gases are finally passed through the engine where ultimate destruction of contaminants occurs.

Effluent concentrations in the water depend on the extent of mass transfer occurring between the water and the gas phase in both the S.A.V.E. and the HFM modules. The removal efficiency is a function of water, air and system parameters, such as flow rates, pressure and temperature, as well as the contaminant's Henry's constant. Henry's constant is the law that when a liquid and a gas are in contact, the mass of the gas that dissolves in a given quantity of liquid is proportional to the pressure of the gas above the liquid. The law holds true only for equilibrium conditions, i.e. when enough time has elapsed so that the quantity of gas is no longer changing

Removal Efficiency Model
To predict removal efficiency in the S.A.V.E. system, we modified the mathematical model developed by Onda (1968) that was used to estimate the mass transfer and removal efficiency.1 According to the model, the total mass transfer that occurs in the S.A.V.E. system can be divided into two phases: liquid and gas phase transfer. The mass transfer coefficients in these two phases are assumed to be related to operating conditions as follows:

Equation 1 and Equation 2







Where
kp
= mass transfer coefficient of phase p (subscript L for Liquid, G for gas) meters per second (m/s)
g
= gravitational acceleration constant, meters per second squared (m/s2)
m
= dynamic viscosity of water or air, centipoise (cP)
LM
= liquid loading rate, kilograms per squared meters per second (kg/m2s)
GM
= gas loading rate (kg/m2s)
r
= density of water or air, kilograms per cubic meter (kg/m3)
D
= diffusivity of solute in water or air (m2/s).

The values of the coefficients C1 to C7 were obtained from experimental studies using the S.A.V.E. system as described below.

The physiochemical properties of contaminants, water and air were obtained from the literature. Gas phase diffusivities of contaminants were calculated according to the Arnold (1930), whereas liquid phase diffusivities are calculated according to Wilke and Chang (1955).2,3 These equations are provided below:

Equation 3





Where
Da
= diffusion coefficient of the pollutant in air (cm2/s)
P
= absolute pressure (atm)
M
= molecular mass (subscript 1 for air, 2 for pollutant), grams per molecule (g/mol)
Vb
= molal volume (subscript 1 for air, 2 for pollutant), (cm3/mol)
S12
= Sutherland constant = 1.47*F*(Tb1Tb2)1/2
where Tb = boiling point, (K)
F
= 1.016-0.0216 Vb2/Vb1

Equation 4





Where
Dw
= diffusion coefficient of the solute in air (cm2/s)
T
= absolute temperature (K)
M
= molecular mass of pollutant (g/mol)
X
= association parameter for the solvent (2.6 for water)
m
= viscosity of water (cP)
Vb
= molal volume of the solute (cm3/mol)

After calculating the mass transfer coefficients for the liquid and gas phase, the overall mass transfer, KLa (s-1) in the system is calculated as:

Equation 5





Where
Hc
= dimensionless Henry?s constant (-)
a
= interfacial unit contact area (m-1).

Equation 6





Where
A, B = coefficients calculated from experimental work
T = temperature in Kelvin.

The temperature dependent values of Hc for MTBE were obtained from Bierwagen and Keller (2001).4 The values of A and B that are used in the model are 6.1 and 2681.8 K-1. A similar approach was used to derive the mass transfer coefficients for the HFM, following Keller et al. (2001).5

Experimental Procedure
The spray aeration vacuum tank has a cylindrical shape, with a diameter of 29 inches and a height of 66 inches. It is made of carbon steel with a special 3M epoxy-coated interior and a powder-coat finish on its exterior. Water flowrate, gas flowrate, pressure and temperature in the tank were controlled by a proprietary microprocessor based controller manufactured by RSI. A 500-gallon tank was used for the experiments, with MTBE added into water and mixed in order to create a homogenous solution. This solution was then run through the S.A.V.E. system at different temperatures. Effluent of the S.A.V.E. system was then passed through the HFM module. For the experiments, water flowrates ranged from 3 gpm to 10 gpm (0.000189 m3/s to 0.00189 m3/s), gas flowrates ranged from 10 scfm to 90 scfm (0.004720 m3/s to 0.075259 m3/s) and the process temperatures ranged from 115 degrees Fahrenheit to 140 degrees Fahrenheit (46 degrees Celsius to 60 degrees Celsius).

In order to measure the system efficiency, water samples were collected at the inlet and outlet ports of the spray tank and the HFM with five minute to 10 minute intervals. The samples were collected into 40-mililiter amber EPA vials and kept in ice for proper preservation. Next, samples were transferred to the laboratory and analyzed using gas chromatography/mass spectrometry Hewlett Packard (HP) 5890/5970 GC/MS for MTBE. Inlet concentrations of MTBE averaged 100 mg/l for the set of experiments conducted. The coefficients in Equations 1-2 were estimated using multi-variate least squares optimization, from experimental data of over 50 experiments.

Results
The coefficient optimization produced the following results (Table 1):


Spray Tank ( S.A.V.E.)

C1

0.5

C2

1.3

C3

-0.3

C4

0.16

C5

1.87

C6

1.25

C7

0.025

Table 1.

For the HFM model, the same approach with the same system was used. However the results for the HFM model show that removal efficiency is not sensitive to gas flowrate. It can be concluded that if the system is operated around 7 scfm to 10 scfm (0.0047 m3/s to 0.0033 m3/s), the removal efficiencies will be essentially the same. The overall mass transfer coefficient for MTBE
is 7.5 x 10-5 s-1. The difference is that the residence time varies as a function of flowrate, such that a higher flowrate results in lower removal efficiency, given the same membrane area.

Application to Predicting Removal
The effluent concentrations of the S.A.V.E. can be predicted using

Equation 7





and the effluent concentrations of the HFM can be described as

Equation 8





Where


Cout
= concentration at the outlet, milligram per liter (mg/L)
Cin
= concentration at the inlet (mg/L)
KLa
= overall mass transfer coefficient (m/s)
L
= length of the membrane (m)
u
= water velocity (m/s)

Finally, the removal efficiency, E (%), of the S.A.V.E. or HFM system can be expressed as:

Equation 9





To predict the removal efficiency of either the S.A.V.E. or HFM in a single pass through either system, one requires only the water and air flowrates, the temperature of each system and the Henry's constant of the contaminant at that temperature. As an example, consider inlet concentrations of MTBE are 100 mg/l, and water flowrate of 3 gpm (0.000189 m3/s), air flowrate of 70 standard cubic feet per minute (scfm) (0.033 m3/s) and a process temperature of 140 degrees Fahrenheit. Using the mass transfer coefficients calculated, the predicted removal efficiency would be 89 percent for MTBE using the S.A.V.E. system. For the current HFM configuration at the same liquid flowrate, 3 gpm (0.000189 m3/s), and air flowrate of 10 scfm (0.0047 m3/s), the removal efficiency would be 79 percent for MTBE.

Using the estimated overall mass transfer coefficients, the removal efficiency of the S.A.V.E. and HFM system can be predicted for different conditions as shown in the above example.

Applied Process Technology's Technologies
Applied Process Technology Inc. has developed two advanced oxidation process (AOP) technologies that effectively destroy recalcitrant groundwater contaminants such as MTBE, TBA and 1,4-dioxane. The first, called HiPOx, is designed for pump-and-treat applications. The other, called Vadozone, is an in-situ treatment solution.

Pump-and-Treat Technology Process Description
The HiPOx process utilizes ozone and hydrogen peroxide to form hydroxyl radicals. Hydroxyl radicals react very rapidly to oxidize volatile organic contaminants (VOCs) into non-hazardous compounds including carbon dioxide (CO2) and water (H2O). This oxidation process occurs in the aqueous phase and does not increase the temperature or pressure of the water because it usually occurs at very low concentrations (<1 percent). HiPOx incorporates high pressure, high-precision distribution of ozone and reagents and high-efficiency mixing. Figure 2 illustrates how higher ozone concentrations (8 percent wt. to 10 percent wt.), higher operating pressure 35 pounds per square inch gauge (psig) 45 psig versus ambient pressure and efficient, in-line mixing are combined to maximize mass transfer and reaction efficiency.



Figure 2. Schematic of HiPOx advanced oxidation system.

HiPOx systems use a series of injection modules similar to that shown in Figure 3. Hydrogen peroxide and ozone are injected at 20 psig to 45 psig. The dosed fluid flows immediately through a mixing section followed by a reaction zone specifically designed for the required residence time. A critical design feature of the distributed-injection approach is the low, local concentrations of ozone and hydrogen peroxide.



Figure 3. Schematic diagram of individual reactor in the HiPOx

  • Case Study: Tustin, Calif.
    A HiPOx pilot test was conducted at a former gas station site on a closed military base in Tustin, Calif. During the test, the HiPOx unit reduced MTBE in the groundwater from approximately 35,400 micrograms per liter (ìg/L to 1 ìg/L. TBA was not detected in the influent and effluent waters. After the pilot test was completed, a full-scale 30-gpm HiPOx unit was installed on the site. During operation of the full-scale system, TBA was present in the influent water. Acetone was also present in the influent water and is formed during treatment as a destruction by-product of MTBE and TBA. Because the destruction of acetone requires a significant amount of ozone, it was determined that a bioreactor would provide a cost-effective method of removing the additional acetone. Therefore, a bioreactor was installed on the effluent of the HiPOx system at this site. Treatment results from the conjunctive systems are shown in Table 2.


MTBE

TBA

Acetone

Influent (ìg/L)

17,000

200

110

HiPOx Effluent (ìg/L)

< 2

< 4

680

Bioreactor Effluent (ìg/L)

< 2

< 4

8.2*

* Trip blank acetone: 5.7 parts per billion.

Table 2. HiPOx system and bioreactor results (Tustin)

The 30-gpm installation has been expanded to operate at flow rates of 50 to 70 gpm.

In-Situ Treatment Technology Process Description
The Vadozone chemical oxidation process, jointly developed with Groundwater & Environmental Services Inc., aggressively destroys dissolved-phase and absorbed-phase contaminants in both groundwater and saturated soils by using three powerful oxidizers together -- ozone, hydrogen peroxide and hydroxyl radicals. This is accomplished by injecting compressed air, oxygen, ozone and hydrogen peroxide at various flows and concentrations into the aquifer or the saturated zone of a treatment site and by optimizing the formation and distribution of hydroxyl radicals in the site?s treatment zone. The reactants are injected into the treatment zone in pre-programmed steps via nested injection wells configured in a site-specific injection matrix.

Ozone is produced on-site from a solid state ozone generator to yield high concentrations of ozone (up to 12 percent by weight) that can be applied at concentrations ranging from 0 to 120,000 mg/L. Oxygen and compressed air are each supplied by on-site generators. Up to 200 pounds/day of 35 percent solution hydrogen peroxide may also be applied.

  • Case Study: Kenton, Del.
    Groundwater beneath a gasoline station became contaminated with MTBE, TBA and BTEX at concentrations as high as 26,572 ìg/L, 27,000 ìg/L, and 26,300 ìg/L, respectively. Contaminant plumes were up to 800 feet long. Pilot tests were performed to compare the potential effectiveness of remedial technologies including soil vapor extraction (SVE), vacuum enhanced groundwater extraction, groundwater extraction and re-injection, air sparging (with and without SVE), total-phase extraction, monitored natural attenuation, bioremediation and chemical oxidation. Data from the tests indicated that chemical oxidation would provide the most cost-effective solution and the shortest life-cycle duration for the project.

GES installed a Vadozone chemical oxidation system and a total of 10 nested injection wells at the site. SVE wells were located in the region of the chemical oxidation wells to control vapor migration. The system?s control panel cycled the system to inject reagents into given points on the site?s injection matrix. After four and a half months of operation, BTEX could not be detected in any of the 45 sample locations on the treatment site. After five months of operation, the site was completely remediated to cleanup levels.

Future Development: Combined Processes
RSI and Applied Process Technology are actively working together to develop a new technology for the treatment of recalcitrant compounds such as MTBE and TBA that represents a combination of their present S.A.V.E. and HiPOx systems in one product. Both companies are optimistic that this joint effort will produce a lower cost/smaller footprint design than their separate counterparts that will provide complete process treatment capability in a single unit.

References

1. Onda, K., Takeuchi, H., Okumoto, Y., 1968. "Mass Transfer Coefficients Between Gas and Liquid Phases in Packed Columns," Chem. Eng. Japan.1(1):56-62.
2. Arnold, J. H., 1930. "Studies in Diffusion," Ind. Eng. Chem. 22:1091-1095.
3. Wilke, C.R., Chang, P., 1955. "Correlation of Diffusion Coefficients in Dilute Solution" AIChe Journal, 2:264-270.
4. Bierwagen, B.G., Keller A.A., 2001. "Measurement of Henry's Law Constant for Methyl tert-Butyl Ether (MTBE) Using Solid-Phase Microextraction," Environ. Toxicol. and Chemistry
, 20(8):1625-1629.
5. Keller, A.A., Bierwagen, B.G., 2001. "Hydrophobic Hollow Fiber Membranes for Treating MTBE-Contaminated Water," Environ. Sci. Tech., 35(9): 1875-1879

e-Sources

This article originally appeared in the 07/01/2003 issue of Environmental Protection.

comments powered by Disqus